By Katherine Rivlin, MD, MSc
Assistant Professor, Obstetrics and Gynecology, The Ohio State University Medical Center, Columbus, OH
SYNOPSIS: In this randomized, noninferiority trial among patients seeking emergency contraception after at least one episode of unprotected intercourse within five days of presentation, the levonorgestrel 52-mg intrauterine device (IUD) was noninferior to the copper T30A IUD at preventing pregnancy one month after IUD insertion. Adverse events between the two groups were similar.
SOURCE: Turok DK, Gero A, Simmons RG, et al. Levonorgestrel vs. copper intrauterine devices for emergency contraception. N Engl J Med 2021;384:335-344.
Although not approved by the Food and Drug Administration (FDA) for use as emergency contraception (EC), clinical evidence supports that the copper intrauterine device (Cu IUD) is a highly effective EC method, with postinsertion pregnancy rates lower than 0.1%.1 This EC failure rate is markedly lower than the rate for the two currently FDA-approved methods of EC available in the United States — oral ulipristal acetate (UA) and oral levonorgestrel (LNG). UA had a 1.8% failure rate and LNG had a 2.6% failure rate in one randomized, non-inferiority trial.2
When choosing birth control methods, patients tend to choose the levonorgestrel intrauterine device (LNG IUD) more frequently than they choose the Cu IUD because of the favorable bleeding profile.3 Because the intrauterine device (IUD) has the added benefit of providing continued birth control when used as EC, the lack of evidence around the LNG IUD for EC has limited patient options.
From 2016 to 2019, six family planning sites in Utah enrolled English- or Spanish-speaking patients between the ages of 18 and 35 years presenting for EC after unprotected sexual intercourse within the last five days. To enroll, patients had to desire IUD initiation and pregnancy prevention for at least one year, have a negative urine pregnancy test (UPT), have a history of regular menses, and know the date of their last menstrual period. The study excluded patients who were breastfeeding, had vaginal bleeding of unknown cause, currently were using a highly effective method of birth control (such as sterilization, an IUD, or a contraceptive implant), had an intrauterine infection in the last three months, had untreated Neisseria gonorrhoeae or Chlamydia trachomatis in the last 30 days, had a copper allergy, had used oral EC in the last five days, or had a known uterine cavity anomaly. Notably, the study team did not exclude patients who reported unprotected intercourse more than five days prior to enrollment, or between six and 14 days before IUD placement.
After a negative UPT, participants were randomized to placement of either the 52-mg LNG IUD or the Cu T380A IUD in a 1:1 ratio. Although participants were unaware of the IUD device they received, providers were aware of the randomization group, given the different appearance and insertion methods of the two IUDs. All inserting providers were nurse practitioners and certified nurse midwives with prior IUD insertion experience.
All participants then were scheduled for one-month follow-up and provided with a home UPT to take the morning prior to their follow-up visit. Participants could report the results of their UPT via text message, through an online survey, or at their follow-up visit. At in-person follow-up, another UPT was performed and the participant was informed of the inserted IUD type. For participants with missing one-month pregnancy data, the study staff called and sent text messages, and if UPT results still were missing, the investigators reviewed the medical records for reports of pregnancy. All participants were contacted again at three months and six months after insertion to assess IUD discontinuation, satisfaction, pain and bleeding outcomes, and adverse events.
Of the 711 participants who underwent randomization, 355 received the LNG IUD and 356 received the Cu IUD. Participants reported an average of 2.1 acts of unprotected intercourse in the five days prior to IUD insertion. At one month following IUD insertion, there was only one pregnancy in the LNG IUD group (0.3%; confidence interval [CI], 0.01-1.7) and no pregnancies in the Cu IUD group (0%; CI, 0.9-1.8). The one pregnancy in the LNG IUD group occurred following an act of unprotected intercourse 48 hours prior to IUD insertion. Pregnancy dating was consistent with conception at that time, confirming an EC failure. The pregnancy ended in a spontaneous abortion at 10 weeks. One Cu IUD expelled, and one participant in the LNG IUD group switched to the Cu IUD at her follow-up visit. Adverse events, such as self-reported bleeding, cramping, and other IUD-associated side effects, were similar between the two groups.
This randomized controlled trial supports that the LNG IUD can provide EC safely and effectively following unprotected intercourse up to five days before insertion. Because the LNG IUD met the noninferiority threshold compared to the Cu IUD, the hormonal IUD joins the Cu IUD as one of our most effective methods of EC, far exceeding the efficacy of the orally available options. These findings are all the more remarkable given that the study team included patients with acts of unprotected intercourse prior to the five-day window, or six to 14 days before IUD placement. Such patients particularly are at high risk of pregnancy, and, therefore, often are excluded from EC studies. Such exclusions are frustrating to clinicians, given how commonly patients presenting either for EC or for a scheduled IUD insertion report acts of unprotected intercourse both in and beyond the preceding five-day window.
The low pregnancy rates in this study pave the way to quick-starting IUDs for such patients, assuming a negative UPT. Providers should remember in such situations to counsel patients on the risks of IUD insertions in the setting of an undiagnosed luteal phase pregnancy. Although such pregnancies are rare, pregnancies that occur in IUD users are more likely to result in miscarriage, ectopic pregnancy, or preterm delivery, or to be complicated by infection.4
Currently, when counseling patients on EC options, providers must keep in mind that both oral LNG and UA may be less effective in patients who are overweight or obese, and that concurrent UA administration and hormonal contraception initiation may reduce the ability of the UA to delay ovulation.5,6 In this study, body mass index did not appear to affect EC efficacy, and with the ongoing contraception provided by the IUD, providers need not worry about when to initiate a new contraceptive method. In their Access to Emergency Contraceptive Committee Opinion, the American College of Obstetricians and Gynecologists (ACOG) reviews barriers to EC access. Although oral LNG is available over the counter, financial, education, and healthcare practice barriers make accessing all EC methods challenging for patients. ACOG recommends counseling that the Cu IUD is the most effective method of EC, and that healthcare providers consider integrating Cu IUD provision into their clinical practices and allow same-day IUD insertion.7 This study supports adding the LNG IUD to the EC toolbox and similarly integrating its use into clinical practice.
In their limitations section, the authors described the selection bias that may have resulted from excluding those patients who declined to be randomized to an IUD type. We do not randomize patients to IUD types in clinical practice, but until now, the absence of the LNG IUD from our EC methods has, in effect, created a similar exclusion. Many of us have counseled patients on EC options, and although some might be interested in an IUD and compelled by its higher EC efficacy, many are turned off by the associated Cu IUD side effects. Therefore, as with all clinical trials, we must pay tribute to those research participants who agreed to randomization, because in doing so they have expanded EC options, allowing our offerings to be both more effective and more patient-centered.
- Cleland K, Zhu H, Goldstuck N, et al. The efficacy of intrauterine devices for emergency contraception: A systematic review of 35 years of experience. Hum Reprod 2012;27:1994-2000.
- Glasier AF, Cameron ST, Fine PM, et al. Ulipristal acetate versus levonorgestrel for emergency contraception: A randomised non-inferiority trial and meta-analysis. Lancet 2010;375:555-562.
- Peipert JF, Zhao Q, Allsworth JE, et al. Continuation and satisfaction of reversible contraception. Obstet Gynecol 2011;117:1105-1113.
- Castano PM, Westhoff CL. Experience with same-day placement of the 52 mg levonorgestrel-releasing intrauterine system. Am J Obstet Gynecol 2020;222:S883.e1-S883.e6.
- Glasier A, Cameron ST, Blithe D, et al. Can we identify women at risk of pregnancy despite using emergency contraception? Data from randomized trials of ulipristal acetate and levonorgestrel. Contraception 2011;84:363-367.
- Brache V, Cochon L, Duijkers IJM, et al. A prospective, randomized, pharmacodynamic study of quick-starting a desogestrel progestin-only pill following ulipristal acetate for emergency contraception. Hum Reprod 2015;30:2785- 2793.
- [No authors listed]. Committee Opinion No. 707 summary: Access to emergency contraception. Obstet Gynecol 2017;130:251-252.